CN117533356B - Intelligent driving assistance system and method - Google Patents
Intelligent driving assistance system and method Download PDFInfo
- Publication number
- CN117533356B CN117533356B CN202410026960.9A CN202410026960A CN117533356B CN 117533356 B CN117533356 B CN 117533356B CN 202410026960 A CN202410026960 A CN 202410026960A CN 117533356 B CN117533356 B CN 117533356B
- Authority
- CN
- China
- Prior art keywords
- data
- module
- traffic
- network
- sensor
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 53
- 230000004927 fusion Effects 0.000 claims abstract description 46
- 230000008569 process Effects 0.000 claims abstract description 21
- 230000005540 biological transmission Effects 0.000 claims description 24
- 238000013507 mapping Methods 0.000 claims description 20
- 238000013528 artificial neural network Methods 0.000 claims description 16
- 125000004122 cyclic group Chemical group 0.000 claims description 12
- 230000006870 function Effects 0.000 claims description 12
- 230000003595 spectral effect Effects 0.000 claims description 11
- 230000009466 transformation Effects 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 10
- 230000000739 chaotic effect Effects 0.000 claims description 9
- 238000005516 engineering process Methods 0.000 claims description 9
- 230000007787 long-term memory Effects 0.000 claims description 8
- 230000006403 short-term memory Effects 0.000 claims description 8
- 230000008859 change Effects 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 6
- 230000015654 memory Effects 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 4
- 230000000306 recurrent effect Effects 0.000 claims description 4
- 238000004445 quantitative analysis Methods 0.000 claims description 3
- 230000007246 mechanism Effects 0.000 abstract description 4
- 238000012731 temporal analysis Methods 0.000 abstract 1
- 238000000700 time series analysis Methods 0.000 abstract 1
- 230000008447 perception Effects 0.000 description 14
- 238000011217 control strategy Methods 0.000 description 5
- 238000005265 energy consumption Methods 0.000 description 4
- 230000004913 activation Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 230000010485 coping Effects 0.000 description 3
- 230000007613 environmental effect Effects 0.000 description 3
- 230000004044 response Effects 0.000 description 3
- 230000035945 sensitivity Effects 0.000 description 3
- 238000012300 Sequence Analysis Methods 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000007405 data analysis Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000002708 enhancing effect Effects 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000002604 ultrasonography Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000003749 cleanliness Effects 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 230000001934 delay Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000036632 reaction speed Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000036962 time dependent Effects 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W60/00—Drive control systems specially adapted for autonomous road vehicles
- B60W60/001—Planning or execution of driving tasks
- B60W60/0011—Planning or execution of driving tasks involving control alternatives for a single driving scenario, e.g. planning several paths to avoid obstacles
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W30/00—Purposes of road vehicle drive control systems not related to the control of a particular sub-unit, e.g. of systems using conjoint control of vehicle sub-units
- B60W30/18—Propelling the vehicle
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
- B60W40/02—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to ambient conditions
- B60W40/04—Traffic conditions
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
- B60W50/0097—Predicting future conditions
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W50/00—Details of control systems for road vehicle drive control not related to the control of a particular sub-unit, e.g. process diagnostic or vehicle driver interfaces
- B60W2050/0001—Details of the control system
- B60W2050/0043—Signal treatments, identification of variables or parameters, parameter estimation or state estimation
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W2720/00—Output or target parameters relating to overall vehicle dynamics
- B60W2720/10—Longitudinal speed
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Transportation (AREA)
- Mechanical Engineering (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention relates to the field of traffic control, in particular to an intelligent driving assistance system and method. Firstly, acquiring original data from each sensor, converting the preprocessed original data into a high-dimensional feature space, analyzing the relation between different data points in the high-dimensional feature space, adjusting the association strength between the data points, and introducing a depth feature fusion network to process and fuse the data from each sensor; then, a traffic prediction model is constructed, future traffic flow changes are predicted, route planning is optimized and vehicle speed is adjusted based on the predicted future traffic flow conditions. The problem that the prior art lacks an effective mechanism to dynamically adjust the association between data is solved; the failure to fully consider real-time traffic data and predictive information results in less optimal route selection and speed adjustment; time series analysis cannot be fully utilized to predict future trends; and the technical problem that the strategy cannot be flexibly adjusted when facing the continuously changing road and traffic conditions.
Description
Technical Field
The invention relates to the field of traffic control, in particular to an intelligent driving assistance system and method.
Background
With the rapid development of automatic driving technology, an intelligent driving assistance system becomes an important research field of the automobile industry. These systems aim to improve driving safety, reduce traffic congestion, improve energy efficiency, and enhance driving experience. However, existing intelligent driving assistance techniques face several challenges, especially in complex and diverse driving environments. For example, how to effectively fuse and analyze large amounts of data from multiple sensors, how to accurately predict traffic flow changes, and how to dynamically optimize route planning and vehicle speed control, etc.
In addition, with the increasing complexity and congestion of urban traffic, higher demands are being placed on intelligent driving assistance systems, in particular in terms of environmental awareness, data processing capacity, real-time reaction speed and system adaptability. These technical challenges motivate the exploration and development of higher-level autopilot technologies aimed at better coping with diverse road conditions and changing traffic environments.
Chinese patent application number: CN202310787019.4, publication date: 2023.09.01 discloses a control method, a system, equipment and a storage medium for intelligent auxiliary driving, and belongs to the technical field of driving auxiliary function safety. The intelligent driving assisting control method comprises the following steps: acquiring initial perception data in a preset area around a target vehicle; the initial perception data comprises first initial perception data and second initial perception data; performing data fusion on the first initial perception data and the second initial perception data to determine target perception data; determining a vehicle control strategy based on the target perception data; decision arbitration is carried out on the vehicle control strategy to obtain a vehicle control instruction; and controlling the target vehicle based on the vehicle control instruction. According to the embodiment of the invention, decision arbitration is carried out on the vehicle control strategy so as to judge some vehicle control strategies deviating from the expected one, so that the vehicle control strategy meeting the requirements is determined, and the corresponding vehicle control instruction is generated, thereby realizing accurate functional response under the unexpected condition and improving the user experience.
However, the above technology has at least the following technical problems: the prior art is not sensitive or accurate enough to respond to environmental changes, and lacks an effective mechanism to dynamically adjust the association between data, so that the accuracy and the robustness of data fusion are insufficient; the failure to fully consider real-time traffic data and predictive information results in route selection and speed adjustment optimization, which may increase travel time and energy consumption; the historical data and time sequence analysis cannot be fully utilized to predict future trends, so that the coping capability of the intelligent driving assistance system in a dynamic environment is limited; and technical problems that the policy may not be flexibly adjusted in the face of constantly changing road and traffic conditions.
Disclosure of Invention
The invention provides an intelligent driving assistance system and method, which solve the problems that the response to environmental change is not sensitive or accurate enough and the correlation between data is dynamically adjusted due to the lack of an effective mechanism in the prior art, so that the accuracy and the robustness of data fusion are insufficient; the failure to fully consider real-time traffic data and predictive information results in route selection and speed adjustment optimization, which may increase travel time and energy consumption; the historical data and time sequence analysis cannot be fully utilized to predict future trends, so that the coping capability of the intelligent driving assistance system in a dynamic environment is limited; and technical problems that the policy may not be flexibly adjusted in the face of constantly changing road and traffic conditions. The intelligent driving auxiliary system is realized, and the safety, the efficiency and the adaptability of driving are obviously improved by comprehensively utilizing the multi-source sensor data and the data processing technology.
The invention provides an intelligent driving assistance system and method, which concretely comprises the following technical scheme:
an intelligent driving assistance system comprising:
the system comprises an acquisition module, a feature mapping module, a data association network building module, an information feedback module, a depth feature fusion module, a traffic prediction module, a route planning module and a vehicle speed adjusting module;
the acquisition module is used for acquiring original data from at least one sensor, wherein the sensor comprises a vehicle-mounted camera, a vehicle-mounted radar, a LiDAR sensor, an ultrasonic sensor, an accelerometer and a gyroscope, and the acquisition module is connected with the feature mapping module in a data transmission mode;
the feature mapping module is used for preprocessing the original data, comprises denoising and normalization, converts the original data into a high-dimensional feature space by using wavelet transformation, and is connected with the data association network building module and the depth feature fusion module in a data transmission mode;
the data association network establishment module is used for utilizing spectral clustering to analyze the relation among different data points in the high-dimensional feature space, constructing an association network among the data points, and is connected with the information feedback module in a data transmission mode;
the information feedback module is used for adjusting the association strength between the data points through a chaotic mapping method, and is connected with the depth characteristic fusion module through a data transmission mode;
the depth feature fusion module is used for processing time series data by using a cyclic neural network and a long-short-term memory network, introducing the depth feature fusion network to process and fuse data from at least one sensor, and connecting the depth feature fusion module with the traffic prediction module in a data transmission mode;
the traffic prediction module is used for constructing a traffic prediction model, predicting future traffic flow change by utilizing the fusion characteristics of the original data and the historical traffic data, and is connected with the route planning module and the vehicle speed adjusting module in a data transmission mode;
the route planning module is used for optimizing route planning by utilizing the output of the traffic prediction module, dynamically evaluating the passing efficiency of each route, and is connected with the vehicle speed adjustment module in a data transmission mode;
the vehicle speed adjusting module is used for adjusting the instant speed of the vehicle according to the predicted future traffic flow state, and is connected with the route planning module in a data transmission mode.
An intelligent driving assistance method, comprising the steps of:
s100: raw data are obtained from each sensor, the raw data are preprocessed, the preprocessed raw data are converted into a high-dimensional feature space through wavelet transformation, the relation between different data points in the high-dimensional feature space is analyzed, the association strength between the data points is adjusted, a cyclic neural network and a long-term and short-term memory network are utilized to process time series data, and a deep feature fusion network is introduced to process and fuse data from each sensor;
s200: and constructing a traffic prediction model, predicting future traffic flow changes, optimizing route planning and adjusting vehicle speed based on the future traffic flow state predicted by the traffic prediction model.
Preferably, the S100 specifically includes:
and (3) analyzing the relation between different data points in the high-dimensional feature space through spectral clustering, and constructing a correlation network between the data points based on the result of the spectral clustering analysis.
Preferably, the step S100 further includes:
the association strength between the data points is adjusted by the chaotic mapping method, so that the association between the data points is dynamically adjusted.
Preferably, the step S100 further includes:
processing sequence data with a recurrent neural network based on the time sequence of the data, the sequence data comprising time sequence data or serialized sensor data; in the recurrent neural network, the hidden state of each moment is calculated based on the current input and the hidden state of the previous moment; and adopting a long-term and short-term memory network to process the output of the cyclic neural network; by combining the cyclic neural network and the long-term and short-term memory network, a deep feature fusion network is introduced.
Preferably, the S200 specifically includes:
constructing a traffic prediction model, specifically, combining the output of the depth feature fusion network and historical traffic data to predict future traffic flow changes; converting the output of the historical traffic data and depth feature fusion network into a graph theory model, and regarding the traffic data as a multidimensional space entity to form a traffic network; and converting the multidimensional space data into weights of edges in the graph theory model by adopting a nonlinear mapping function.
Preferably, the step S200 further includes:
and predicting the future traffic flow state by using a dynamic optimization technology and a nonlinear system theory, and introducing a traffic flow state prediction formula.
Preferably, the step S200 further includes:
optimizing route planning by utilizing future traffic flow states predicted by a traffic prediction model, and dynamically evaluating the traffic efficiency of each route; an efficiency score is calculated for each route, and an optimal route is determined by quantitative analysis.
Preferably, the step S200 further includes:
the instantaneous speed of the vehicle is adjusted according to the predicted future traffic flow conditions to cope with the predicted traffic conditions of the road section ahead.
The technical scheme of the invention has the beneficial effects that:
1. by converting the original data from different sensors into a high-dimensional feature space and applying wavelet transformation and spectral clustering technology, the fine difference and complex mode of the data can be more effectively captured, and the comprehensive and deep data analysis improves the accuracy and the fineness of the vehicle environment perception; the association strength between the data points is adjusted through the chaotic mapping method, so that complex and variable data modes can be dynamically adapted, the accuracy of data fusion is enhanced, and the robustness of the system to uncertainty and noise is improved;
2. by constructing a traffic prediction model, future traffic flow changes can be predicted, route planning and vehicle speed control are effectively optimized, travel time is reduced, congestion is avoided, and running efficiency is improved; by intelligently adjusting the vehicle speed, not only is the driving safety improved, but also the energy efficiency is improved by reducing unnecessary acceleration and deceleration, and the method is particularly important for reducing the energy consumption and the emission; the time sequence data is processed by the cyclic neural network and the long-term and short-term memory network, so that the change mode of the data along with time can be captured, and more accurate and comprehensive environment perception information is provided for the intelligent driving assistance system.
Drawings
FIG. 1 is a block diagram of an intelligent driving assistance system according to an embodiment of the present invention;
fig. 2 is a flowchart of an intelligent driving assistance method according to an embodiment of the present invention.
Detailed Description
In order to further illustrate the technical means and effects adopted by the present invention to achieve the preset purpose, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of an intelligent driving assistance system and method provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, there is shown a block diagram of an intelligent driving assistance system according to an embodiment of the present invention, the system includes:
the system comprises an acquisition module, a feature mapping module, a data association network building module, an information feedback module, a depth feature fusion module, a traffic prediction module, a route planning module and a vehicle speed adjusting module;
the acquisition module is used for acquiring original data from various sensors, wherein the various sensors comprise a vehicle-mounted camera, a vehicle-mounted radar (such as millimeter wave radar), a LiDAR sensor, an ultrasonic sensor, an accelerometer, a gyroscope and the like, and the acquisition module is connected with the feature mapping module in a data transmission mode;
the feature mapping module is used for preprocessing the original data, comprises denoising and normalization, converts the original data into a high-dimensional feature space by using wavelet transformation, and is connected with the data association network building module and the depth feature fusion module in a data transmission mode;
the data association network establishment module is used for utilizing spectral clustering to analyze the relation among different data points in the high-dimensional feature space, constructing an association network among the data points, and is connected with the information feedback module in a data transmission mode;
the information feedback module is used for adjusting the association strength between data points through a chaotic mapping method, dynamically adapting to complex and changed data modes, enhancing the accuracy and the robustness of data fusion, and being connected with the depth feature fusion module through a data transmission mode;
the depth feature fusion module is used for processing time sequence data by using a cyclic neural network (RNN) and a long-short-term memory network (LSTM) to capture the mode of data changing along with time, so as to realize feature fusion, and the depth feature fusion module is connected with the traffic prediction module in a data transmission mode;
the traffic prediction module is used for constructing a traffic prediction model, predicting future traffic flow change by utilizing fusion characteristics and historical traffic data, and is connected with the route planning module and the vehicle speed adjusting module in a data transmission mode;
the route planning module is used for optimizing route planning by utilizing the output of the traffic prediction module, dynamically evaluating the passing efficiency of each route, and is connected with the vehicle speed adjustment module in a data transmission mode;
the vehicle speed adjusting module is used for adjusting the instant speed of the vehicle according to the predicted future traffic flow state so as to adapt to the expected traffic condition, and is connected with the route planning module in a data transmission mode.
Referring to fig. 2, a flowchart of an intelligent driving assistance method according to an embodiment of the present invention is shown, the method includes the following steps:
s100: raw data are obtained from each sensor, the raw data are preprocessed, the preprocessed raw data are converted into a high-dimensional feature space through wavelet transformation, the relation between different data points in the high-dimensional feature space is analyzed, the association strength between the data points is adjusted, a cyclic neural network and a long-term and short-term memory network are utilized to process time series data, and a deep feature fusion network is introduced to process and fuse data from each sensor;
the acquisition module acquires raw data from each sensor, including image or video data from an onboard camera, distance measurements from onboard radar (e.g., millimeter wave radar), high-precision distance and shape information provided by LiDAR sensors, ultrasound data from ultrasound sensors, data from accelerometers and gyroscopes, and so forth.
The feature mapping module performs preprocessing, including denoising and normalization, on the raw data collected by each sensor, so as to ensure the cleanliness and consistency of the data and eliminate the difference between different devices. For example, both image data from an onboard camera and range measurements from an onboard radar are converted to a uniform format and scale. To better capture local features and details of the data, wavelet transforms are employed to transform the preprocessed raw data into a high-dimensional feature space in order to more effectively capture subtle differences and complex patterns of the data. Wavelet transformation formula:
，
wherein,representing a high-dimensional characteristic representation of the s-th sensor data after wavelet transformation; />Is the original data collected by the s-th sensor; />Is the wavelet basis function corresponding to the data of the s-th sensor at the k-th frequency level; />Is the wavelet coefficient, i.e. the characteristic intensity of the data of the s-th sensor at the k-th frequency level.
The data association network building module utilizes spectral clustering to analyze the relation between different data points in the high-dimensional feature space, identifies the natural distribution of the data points in the high-dimensional space, and helps to understand the mutual association between different sensor data; and constructing a correlation network among the data points through the result of spectral cluster analysis. The calculation formula of the spectral clustering is as follows:
，
wherein,representing the distances of the ith and jth data points in the high-dimensional feature space after spectral cluster transformation; />And->The characteristic value and the characteristic vector of the Laplace matrix respectively representing the ith data point are used for measuring the distance between the data points in the characteristic space; />Representing the transpose.
The information feedback module further adjusts the association strength between the data points through a chaotic mapping method so as to dynamically adjust the association between the data points, and can adapt to complex and changing data modes, thereby enhancing the accuracy and the robustness of data fusion. The chaotic map updating formula is as follows:
，
wherein,representing the correlation strength between the ith and jth data points at the nth iteration; />Is a parameter of the chaotic system for controlling the update rate and pattern of the correlation strength between data points.
Considering the time sequence of the data, the depth feature fusion module processes the time sequence data by using a cyclic neural network (RNN) and a long-short-term memory network (LSTM), captures the mode of the data changing along with time and realizes the fusion of the features.
In particular, RNNs are used to process sequence data, such as time series data or serialized sensor data. The RNN is able to remember the information of the previous moment and use it for data processing at the current moment, which makes the RNN particularly suitable for processing time-dependent data. In the RNN, the hidden state at each instant is calculated based on the current input and the hidden state at the previous instant. The basic formula of RNN can be expressed as:
，
wherein,representing the hidden state of the RNN at the point in time t; />Is the weight of the input layer to the hidden layer;indicating that +.>And->Is a function of (2); />Is the weight from hidden state to hidden state; />Is the bias of the input layer to the hidden layer; />Is a bias term for the hidden state; />Is an activation function for introducing nonlinearities.
The LSTM network then processes the RNN output to further refine and strengthen the long-term dependencies in the time series data. LSTM is capable of better controlling the retention and forgetting of information by introducing gating mechanisms (including forgetting gate, input gate and output gate), thereby effectively handling longerIs a sequence data of (a) in a sequence data set. The core of the LSTM unit is composed of several parts including forgetting gateInput door->Output door->And cell state->. The formulas for these components are as follows:
，
，
，
，
，
，
wherein,、/>weights and biases representing forget gates; />And->Representing the weight and bias of the input gate;and->Weights and biases representing cell states; />And->Representing the weights and offsets of the output gates; />Is a candidate cell state; />Is a sigmoid activation function for gate control; />Is another activation function for creating new candidate values;is the hidden state of the LSTM at time point t-1.
By combining the characteristics of RNN and LSTM, a depth feature fusion network is introduced, and the depth feature fusion network can effectively process and fuse multi-source data with time dependency, so that more accurate and comprehensive environment perception information is provided in the intelligent driving assistance system.
，
Wherein,is the output of the depth feature fusion network, +.>And->The weight and bias parameters of the depth feature fusion network, respectively. By the method, data from a plurality of sensors are effectively processed and fused, highly accurate and reliable environment perception information is provided, and the performance and safety of the intelligent driving assistance system are greatly improved.
S200: and constructing a traffic prediction model, predicting future traffic flow changes, optimizing route planning and adjusting vehicle speed based on the future traffic flow state predicted by the traffic prediction model.
In order to quickly and effectively process and respond to dynamic traffic conditions, a traffic prediction module constructs a traffic prediction model, and utilizes the fusion characteristics of original data and historical traffic data to predict future traffic flow changes. The traffic flow is accurately predicted and the vehicle energy management is optimized by adopting a method combining nonlinear system theory, graph theory, dynamic optimization technology, multi-stage decision process and random process.
Specifically, historical traffic data is used forAnd the output of the depth feature fusion network +.>Conversion to graph theory model->And regarding the traffic data as a multidimensional space entity to form a traffic network. In a multidimensional space, each dimension represents a different attribute of the traffic flow, such as flow, speed, and accident frequency. And converting the data in the multidimensional space into the weights of the edges in the graph theory model by adopting a nonlinear mapping function based on topology and complex network theory. This conversion process can more accurately capture the structural characteristics and dynamic changes of the traffic network. The formula is expressed as follows:
，
wherein,representing a graph theory model; />Represents an intersection (vertex of the graph); />Representing the road (edge of the graph); />A weight representing the i-th dimension; />Is corresponding dimension->For converting traffic data into weights for edges in the graph theory model.
Predicting future traffic flow conditions using dynamic optimization techniques and nonlinear system theoryThe adopted traffic flow state prediction formula comprehensively considers each node and each edge in the traffic network, and the sensitivity and the response of the model are adjusted through parameters. The traffic flow state prediction formula is as follows:
，
wherein,and->Is a model parameter of the jth road segment; />The sensitivity of the model to the traffic flow of the jth road section is adjusted; />The method comprises the steps of controlling the influence of a distance function on traffic flow prediction; />Is the vertex->And (2) He Ji->And a distance function between the two nodes represents the connection strength of different nodes in the traffic network.
Future traffic flow state predicted by route planning module through traffic prediction modelOptimizing route planning, dynamically evaluating traffic efficiency for each route, including a comprehensive consideration of predicted traffic density, possible delays, and any potential route disturbances for each road segment. An 'efficiency score' is calculated for each route that integrates time, distance and predicted traffic conditions, and an optimal route is determined by quantitative analysis, reducing total travel time and avoiding traffic congestion. The specific efficiency score calculation formula is:
，
wherein,representing the efficiency score of a route, +.>The predicted traffic time of the road section z considers traffic flow changes of different road sections; />Is an adjustment parameter considering distance factors and is used for balancing the influence of the far and near road sections; />Is the length of the road section z, and directly influences the travel time; />Is based on->The traffic density function of the road section z of the road section reflects the congestion degree of different road sections; />Representing the total number of road segments.
The vehicle speed adjusting module adjusts the vehicle speed according to the predicted future traffic flow stateThe instantaneous speed of the vehicle is adjusted to account for the predicted traffic conditions of the road segment ahead. For example, slow down early when congestion is predicted, or speed up moderately when traffic is clear. The formula is defined as:
，
，
wherein,the adjusted speed determines the actual running speed of the vehicle; />The current speed is the current running state of the vehicle; />Is a speed adjustment sensitivity parameter, which controls the amplitude of speed adjustment;is based onPredicted future traffic flow state->And ideal driving speed +.>For fine tuning the speed to achieve optimal driving efficiency and safety; />Representing road speed limit; />Representing an energy efficiency optimal speed; />Indicating a safe speed.
Through the calculation, the route planning and the vehicle speed control are effectively optimized in a complex and constantly changing traffic environment. The running efficiency and the safety are improved, and the value and the feasibility of the invention in practical application are ensured.
In summary, an intelligent driving assistance system and method are completed.
According to the embodiment of the invention, the original data from different sensors are converted into the high-dimensional feature space, and the wavelet transformation and the spectral clustering technology are applied, so that the subtle difference and the complex mode of the data can be more effectively captured, and the comprehensive and deep data analysis improves the accuracy and the fineness of the vehicle environment perception; the association strength between the data points is adjusted through the chaotic mapping method, so that complex and variable data modes can be dynamically adapted, the accuracy of data fusion is enhanced, and the robustness of the system to uncertainty and noise is improved;
according to the embodiment of the invention, the traffic prediction model is constructed, so that future traffic flow changes can be predicted, route planning and vehicle speed control are effectively optimized, travel time is reduced, congestion is avoided, and running efficiency is improved; by intelligently adjusting the vehicle speed, not only is the driving safety improved, but also the energy efficiency is improved by reducing unnecessary acceleration and deceleration, and the method is particularly important for reducing the energy consumption and the emission; the RNN and LSTM are utilized to process time sequence data, so that the change mode of the data along with time can be captured, and more accurate and comprehensive environment perception information is provided for an intelligent driving auxiliary system.
The sequence of the embodiments of the invention is merely for description and does not represent the advantages or disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and the same or similar parts of each embodiment are referred to each other, and each embodiment mainly describes differences from other embodiments.
The above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention, and are intended to be included in the scope of the present invention.
Claims (6)
1. An intelligent driving assistance system, comprising:
the system comprises an acquisition module, a feature mapping module, a data association network building module, an information feedback module, a depth feature fusion module, a traffic prediction module, a route planning module and a vehicle speed adjusting module;
the acquisition module is used for acquiring original data from at least one sensor, wherein the sensor comprises a vehicle-mounted camera, a vehicle-mounted radar, a LiDAR sensor, an ultrasonic sensor, an accelerometer and a gyroscope, and the acquisition module is connected with the feature mapping module in a data transmission mode;
the feature mapping module is used for preprocessing the original data, comprises denoising and normalization, converts the original data into a high-dimensional feature space by using wavelet transformation, and is connected with the data association network building module and the depth feature fusion module in a data transmission mode;
the data association network establishment module is used for utilizing spectral clustering to analyze the relation among different data points in the high-dimensional feature space, constructing an association network among the data points, and is connected with the information feedback module in a data transmission mode;
the information feedback module is used for adjusting the association strength between the data points through a chaotic mapping method, and is connected with the depth characteristic fusion module through a data transmission mode;
the depth feature fusion module is used for processing time series data by using a cyclic neural network and a long-short-term memory network, introducing the depth feature fusion network to process and fuse data from at least one sensor, and connecting the depth feature fusion module with the traffic prediction module in a data transmission mode;
the traffic prediction module is used for constructing a traffic prediction model, predicting future traffic flow change by utilizing the fusion characteristics of the original data and the historical traffic data, and is connected with the route planning module and the vehicle speed adjusting module in a data transmission mode;
the route planning module is used for optimizing route planning by utilizing the output of the traffic prediction module, dynamically evaluating the passing efficiency of each route, and is connected with the vehicle speed adjustment module in a data transmission mode;
the vehicle speed adjusting module is used for adjusting the instant speed of the vehicle according to the predicted future traffic flow state, and is connected with the route planning module in a data transmission mode.
2. An intelligent driving assistance method applied to the intelligent driving assistance system according to claim 1, comprising the steps of:
s100: raw data are obtained from each sensor, the raw data are preprocessed, the preprocessed raw data are converted into a high-dimensional feature space through wavelet transformation, the relation between different data points in the high-dimensional feature space is analyzed, the association strength between the data points is adjusted, a cyclic neural network and a long-term and short-term memory network are utilized to process time series data, and a deep feature fusion network is introduced to process and fuse data from each sensor;
s200: and constructing a traffic prediction model, predicting future traffic flow changes, optimizing route planning and adjusting vehicle speed based on the future traffic flow state predicted by the traffic prediction model.
3. The intelligent driving assistance method according to claim 2, characterized in that the S100 further comprises:
processing sequence data with a recurrent neural network based on the time sequence of the data, the sequence data comprising time sequence data or serialized sensor data; in the recurrent neural network, the hidden state of each moment is calculated based on the current input and the hidden state of the previous moment; and adopting a long-term and short-term memory network to process the output of the cyclic neural network; by combining the cyclic neural network and the long-term and short-term memory network, a deep feature fusion network is introduced.
4. The intelligent driving assistance method according to claim 2, wherein S200 specifically includes:
constructing a traffic prediction model, specifically, combining the output of the depth feature fusion network and historical traffic data to predict future traffic flow changes; converting the output of the historical traffic data and depth feature fusion network into a graph theory model, and regarding the traffic data as a multidimensional space entity to form a traffic network; and converting the multidimensional space data into weights of edges in the graph theory model by adopting a nonlinear mapping function.
5. The intelligent driving assistance method according to claim 4, characterized in that said S200 further comprises:
and predicting the future traffic flow state by using a dynamic optimization technology and a nonlinear system theory, and introducing a traffic flow state prediction formula.
6. The intelligent driving assistance method according to claim 5, characterized in that S200 further comprises:
optimizing route planning by utilizing future traffic flow states predicted by a traffic prediction model, and dynamically evaluating the traffic efficiency of each route; an efficiency score is calculated for each route, and an optimal route is determined by quantitative analysis.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410026960.9A CN117533356B (en) | 2024-01-09 | 2024-01-09 | Intelligent driving assistance system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410026960.9A CN117533356B (en) | 2024-01-09 | 2024-01-09 | Intelligent driving assistance system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117533356A CN117533356A (en) | 2024-02-09 |
| CN117533356B true CN117533356B (en) | 2024-03-29 |
Family
ID=89794171
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410026960.9A Active CN117533356B (en) | 2024-01-09 | 2024-01-09 | Intelligent driving assistance system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117533356B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117787838B (en) * | 2024-02-27 | 2024-05-28 | 广州一链通互联网科技有限公司 | Logistics digital management system and method based on AI large model |
| CN117975696A (en) * | 2024-03-28 | 2024-05-03 | 南京邦固消防科技有限公司 | Linkage type fire alarm control system and method |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109118014A (en) * | 2018-08-30 | 2019-01-01 | 浙江工业大学 | A kind of traffic flow speed prediction technique based on time recurrent neural network |
| CN110021165A (en) * | 2019-03-18 | 2019-07-16 | 浙江工业大学 | A kind of traffic flow forecasting method based on Autoencoder-LSTM Fusion Model |
| CN110634294A (en) * | 2019-09-26 | 2019-12-31 | 刘秀萍 | Time-space traffic flow prediction method driven by reinforced hierarchical learning |
| CN113763700A (en) * | 2021-04-26 | 2021-12-07 | 腾讯云计算（北京）有限责任公司 | Information processing method, information processing device, computer equipment and storage medium |
| CN114463977A (en) * | 2022-02-10 | 2022-05-10 | 北京工业大学 | Path planning method based on vehicle-road collaborative multi-source data fusion traffic flow prediction |
| CN114783183A (en) * | 2022-04-15 | 2022-07-22 | 中远海运科技股份有限公司 | Monitoring method and system based on traffic situation algorithm |
| WO2022241802A1 (en) * | 2021-05-19 | 2022-11-24 | 广州广电运通金融电子股份有限公司 | Short-term traffic flow prediction method under complex road network, storage medium, and system |
| CN115713856A (en) * | 2022-10-19 | 2023-02-24 | 东南大学 | Vehicle path planning method based on traffic flow prediction and actual road conditions |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112561146B (en) * | 2020-12-08 | 2023-04-18 | 哈尔滨工程大学 | Large-scale real-time traffic flow prediction method based on fuzzy logic and depth LSTM |
-
2024
- 2024-01-09 CN CN202410026960.9A patent/CN117533356B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109118014A (en) * | 2018-08-30 | 2019-01-01 | 浙江工业大学 | A kind of traffic flow speed prediction technique based on time recurrent neural network |
| CN110021165A (en) * | 2019-03-18 | 2019-07-16 | 浙江工业大学 | A kind of traffic flow forecasting method based on Autoencoder-LSTM Fusion Model |
| CN110634294A (en) * | 2019-09-26 | 2019-12-31 | 刘秀萍 | Time-space traffic flow prediction method driven by reinforced hierarchical learning |
| CN113763700A (en) * | 2021-04-26 | 2021-12-07 | 腾讯云计算（北京）有限责任公司 | Information processing method, information processing device, computer equipment and storage medium |
| WO2022241802A1 (en) * | 2021-05-19 | 2022-11-24 | 广州广电运通金融电子股份有限公司 | Short-term traffic flow prediction method under complex road network, storage medium, and system |
| CN114463977A (en) * | 2022-02-10 | 2022-05-10 | 北京工业大学 | Path planning method based on vehicle-road collaborative multi-source data fusion traffic flow prediction |
| CN114783183A (en) * | 2022-04-15 | 2022-07-22 | 中远海运科技股份有限公司 | Monitoring method and system based on traffic situation algorithm |
| CN115713856A (en) * | 2022-10-19 | 2023-02-24 | 东南大学 | Vehicle path planning method based on traffic flow prediction and actual road conditions |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117533356A (en) | 2024-02-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117533356B (en) | Intelligent driving assistance system and method | |
| CN108091135B (en) | Parking position multi-step prediction method based on optimized wavelet neural network | |
| CN113538910B (en) | Self-adaptive full-chain urban area network signal control optimization method | |
| Dou et al. | A hybrid CEEMD-GMM scheme for enhancing the detection of traffic flow on highways | |
| Yu et al. | A special event-based K-nearest neighbor model for short-term traffic state prediction | |
| WO2019071909A1 (en) | Automatic driving system and method based on relative-entropy deep inverse reinforcement learning | |
| CN113779267B (en) | On-board intelligent task decision method based on intention | |
| Gao et al. | Short-term traffic speed forecasting using a deep learning method based on multitemporal traffic flow volume | |
| CN115675098B (en) | Intelligent network-connected automobile kinetic energy recovery system, method, equipment and storage medium based on user characteristic difference | |
| CN112258850A (en) | Edge side multi-sensor data fusion system of vehicle-road cooperative system | |
| Qi et al. | A deep learning approach for long-term traffic flow prediction with multifactor fusion using spatiotemporal graph convolutional network | |
| Song et al. | Collaborative processing and data optimization of environmental perception technologies for autonomous vehicles | |
| Tian et al. | RETRACTED ARTICLE: Research on emergency vehicle routing planning based on short-term traffic flow prediction | |
| Tang et al. | Prediction about the vessel’s heave motion under different sea states based on hybrid PSO_ARMA model | |
| Shi et al. | Short-term vehicle speed prediction by time series neural network in high altitude areas | |
| Chong et al. | A novel neuro-cognitive approach to modeling traffic control and flow based on fuzzy neural techniques | |
| CN115034478B (en) | Traffic flow prediction method based on field self-adaption and knowledge migration | |
| CN111160089A (en) | Trajectory prediction system and method based on different vehicle types | |
| CN116978218A (en) | Urban traffic flow prediction method, system and equipment based on generation countermeasure network | |
| CN116384695A (en) | Unmanned aerial vehicle operation monitoring method and system based on independent overruling and combined overruling | |
| CN114386672B (en) | Environment big data Internet of things intelligent detection system | |
| Tang et al. | Actively learning Gaussian process dynamical systems through global and local explorations | |
| Xiang et al. | Adaptive traffic signal control of bottleneck subzone based on grey qualitative reinforcement learning algorithm | |
| CN113689694A (en) | Traffic flow prediction method, device, equipment and readable storage medium | |
| Zhao et al. | Application of data fusion based on clustering-neural network for ETC gantry flow capacity correction |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |